<script language="Javascript">
function openNewWindow(fileName,windowName,theWidth,theHeight) {
if (windowName == "newMessageWindow")
{
//generate random window ID
windowName = new String(Math.round(Math.random() * 100000));
}
window.open(fileName,windowName,"toolbar=0,location=0,directories=0,status=1,menubar=1,scrollbars=1,resizable=1,width="+theWidth+",height="+theHeight)
}
</script>
<!DOCTYPE HTML PUBLIC \"-//IETF//DTD HTML 3.2//EN\">
<HTML>
<HEAD>
<TITLE>Re: RE2: HFS+ a bit Progress</TITLE>
</HEAD>
<SCRIPT language=javascript>
function DoCommand(szCommand) {
if (szCommand == "reply") {
window.location="commands.asp?command=reply&obj=0000000093B183274AFBD311AF1F005004458BD10700FD5B31C730D6D311AF12005004458BD10000000049C80000F14441D40E5AD3118886005004607AD100000017A48B0000";
}
else if (szCommand == "delete"){
window.location="commands.asp?command=delete&obj=0000000093B183274AFBD311AF1F005004458BD10700FD5B31C730D6D311AF12005004458BD10000000049C80000F14441D40E5AD3118886005004607AD100000017A48B0000";
}
else if (szCommand == "replyall"){
window.location="commands.asp?command=replyall&obj=0000000093B183274AFBD311AF1F005004458BD10700FD5B31C730D6D311AF12005004458BD10000000049C80000F14441D40E5AD3118886005004607AD100000017A48B0000";
}
else if (szCommand == "forward"){
window.location="commands.asp?command=forward&obj=0000000093B183274AFBD311AF1F005004458BD10700FD5B31C730D6D311AF12005004458BD10000000049C80000F14441D40E5AD3118886005004607AD100000017A48B0000";
}
else if (szCommand == "replytofolder"){
window.location="commands.asp?command=replytofolder&obj=0000000093B183274AFBD311AF1F005004458BD10700FD5B31C730D6D311AF12005004458BD10000000049C80000F14441D40E5AD3118886005004607AD100000017A48B0000";
}
else if (szCommand == "close"){
window.close();
}
else if (szCommand == "next"){
window.location="/exchange/item.asp?action=next";
}
else if (szCommand == "previous"){
window.location="/exchange/item.asp?action=prev";
}
}
</Script>
<BODY TEXT=000000 BGCOLOR=#c0c0c0 TOPMARGIN=4 LEFTMARGIN=4 text=000000 link=000000 vlink=000000 alink=000000>
<!--- begin toolbar --------->
<form>
<TABLE BORDER=0 bgcolor=#c0c0c0 CELLSPACING=0 CELLPADDING=0 WIDTH="100%">
<tr>
<td width=100%>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
<A href="JavaScript:DoCommand('reply')">
<IMG SRC="/exchange/forms/reply.gif" Alt="Reply to sender" align="middle" border=0></a>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
<A href="JavaScript:DoCommand('replyall')">
<IMG SRC="/exchange/forms/replyall.gif" Alt="Reply to all" align="middle" border=0></a>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
<A HREF="JavaScript:DoCommand('replytofolder')">
<IMG SRC="/exchange/forms/ReplyFld.gif" alt='Reply to folder' align="middle" WIDTH=24 HEIGHT=24 BORDER=0></a>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
<A href="JavaScript:DoCommand('forward')">
<IMG SRC="/exchange/forms/forward.gif" Alt="Forward" align="middle" border=0 height=24 width=24></a>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
<a href="JavaScript:parent.openNewWindow('/exchange/movcpy/root.asp?msgid=0000000093B183274AFBD311AF1F005004458BD10700FD5B31C730D6D311AF12005004458BD10000000049C80000F14441D40E5AD3118886005004607AD100000017A48B0000&folderid=0000000093B183274AFBD311AF1F005004458BD10100FD5B31C730D6D311AF12005004458BD10000000049C80000&process=1','newMessageWindow',400,400)">
<IMG SRC="/exchange/forms/movcpy.gif" align="middle" Alt="Move/Copy" border=0 height=24 width=24></a>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
<A href="JavaScript:DoCommand('delete')">
<IMG SRC="/exchange/forms/delmark.gif" Alt='Delete' align="middle" border=0 height=24 width=24></a>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
<A href="JavaScript:DoCommand('previous')">
<IMG SRC="/exchange/forms/prevmsg.gif" Alt="Read previous item" align="middle" border=0 width=24 height=24></a>
<A href="JavaScript:DoCommand('next')">
<IMG SRC="/exchange/forms/nextmsg.gif" Alt='Read next item' align="middle" border=0 width=24 height=24></a>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
<A href="JAVASCRIPT:openNewWindow('/exchange/help/READMSG.HTM','inlineHelpWindow',600,400)">
<IMG SRC="/exchange/images/help.gif" alt="Get help information on the current window" align="middle" border=0 height=20 width=20></a>
<img src="/exchange/images/divider.gif" width=4 height=24 align="middle">
</td>
<td align="right">
<input type="button" value="Close" onClick="DoCommand('close')">
</td>
</tr>
</table>
</form>
<!--<br>-->
<!--- begin header --->
<TABLE border=0 cellpadding=3 cellspacing=0 width="100%" bgcolor=#c0c0c0>
<TR>
<TD width=10% >
<font size=2><b>From:</b></font>
</TD>
<TD colspan=3><font size=2>
flar@allandria.com
[SMTP:flar@allandria.com]
</font></TD>
</TR>
<TR>
<TD>
<font size=2><b>To:</b></font>
</TD>
<TD colspan=3>
<font size=2>
khalfmann@libra.de
</font>
</TD>
</TR>
<TR>
<TD><font size=2><b>Cc:</b></font>
</TD>
<TD colspan=3><font size=2>
</font>
</TD>
</TR>
<TR>
<TD> </TD>
</TR>
<TR>
<TD>
<font size=2><b>Subject:</b></font>
</TD>
<TD colspan=3>
<font size=2>
Re: RE2: HFS+ a bit Progress
</font>
</TD>
</TR>
<TR>
<TD>
<font size=2><nobr><b>Sent:</b></nobr>
</TD>
<TD width=62% ><font size=2><nobr>
8/19/00 4:38 AM </nobr></font><br>
</TD>
<TD align=right>
<font size=2><nobr><b>Importance:</b></nobr>
</TD>
<TD align=right><font size=2><nobr>
Normal </nobr>
</font>
</TD>
</TR>
</TABLE>
<!--- message text ----->
<TABLE width="100%" border=1 cellspacing=2 cellpadding=6>
<TR>
<TD BGCOLOR=white width=100% height=300 valign=top colspan=2 >
<P><FONT SIZE=2>Halfmann, Klaus wrote:</FONT>
<BR><FONT SIZE=2>> Uhm, I found this one, I thougt that btree_head.reserved3</FONT>
<BR><FONT SIZE=2>> Would be [3] (reserverd_3_) but it actually is [16].</FONT>
</P>
<P><FONT SIZE=2>Ah, yes. The names reserved[23] just mean that they're the second</FONT>
<BR><FONT SIZE=2>or third reserved block. I don't really like the struct member</FONT>
<BR><FONT SIZE=2>names as given in official Apple definitions.</FONT>
</P>
<P><FONT SIZE=2>> I still can not find these offset, can you tell me where in </FONT>
<BR><FONT SIZE=2>> dump1.txt I can find them ?</FONT>
</P>
<P><FONT SIZE=2>004dbf0: 0000 0000 0000 0000 0ff8 00f8 0078 000e .............x..</FONT>
</P>
<P><FONT SIZE=2>Here's the one for the header node in the catalog tree. The three</FONT>
<BR><FONT SIZE=2>records are at the following offsets:</FONT>
</P>
<P><FONT SIZE=2>r1: 000e (14)</FONT>
<BR><FONT SIZE=2>r2: 0078 (120)</FONT>
<BR><FONT SIZE=2>r3: 00f8 (248)</FONT>
</P>
<P><FONT SIZE=2>Also, the free space begins at 0ff8.</FONT>
</P>
<P><FONT SIZE=2>> Yes , I just started to grasp whats up with the extends overflow</FONT>
<BR><FONT SIZE=2>> file. I intended to add this functionality later. (Ill come</FONT>
<BR><FONT SIZE=2>> away without for some time I assume :)</FONT>
</P>
<P><FONT SIZE=2>Not a huge deal. Only a heavily fragmented filesystem ever uses</FONT>
<BR><FONT SIZE=2>this. I don't have any HFS+ filesystems where this isn't empty.</FONT>
</P>
<P><FONT SIZE=2>> I once wrote Apple Progs only (sigh) and know very well how</FONT>
<BR><FONT SIZE=2>> to do that :). Here are some suggestions Ill try when you dont</FONT>
<BR><FONT SIZE=2>> mind 'LiH0', 'Linu', 'LiH+' </FONT>
</P>
<P><FONT SIZE=2>Sounds fine. I guess I'll have to get some for the kernel as well,</FONT>
<BR><FONT SIZE=2>whenever I actually make it write to the disk.</FONT>
</P>
<P><FONT SIZE=2>> Hmm, Im on irc from about 08:00-17:00 EST (European Summer time) </FONT>
<BR><FONT SIZE=2>> at workdays (Mo-Fr). _Sometimes_ Im online via modem around </FONT>
<BR><FONT SIZE=2>> 18:00 - 20:00 EST (but only short, is somehwat costly ...)</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> Can you be online at work ? </FONT>
</P>
<P><FONT SIZE=2>Yes, I have a very good net connection at work. I would be able</FONT>
<BR><FONT SIZE=2>to catch the last couple hours of the time you are on that way.</FONT>
</P>
<P><FONT SIZE=2>> Do you receive mails online ? If yes I could send you</FONT>
<BR><FONT SIZE=2>> a mail and upon a direct answer we could go online simultaneously.</FONT>
</P>
<P><FONT SIZE=2>I don't generally look at my personal email while I'm at work, but</FONT>
<BR><FONT SIZE=2>I keep an eye on my work email. (bboyer@yahoo-inc.com)</FONT>
</P>
<P><FONT SIZE=2>I also am almost always logged in with Yahoo! messenger during the</FONT>
<BR><FONT SIZE=2>day as well. I even have a linuxppc build of the client. (It isn't</FONT>
<BR><FONT SIZE=2>on the official download site yet, but it should get added soon.)</FONT>
<BR><FONT SIZE=2><A HREF="http://messenger.yahoo.com/" TARGET="_blank">http://messenger.yahoo.com/</A></FONT>
</P>
<P><FONT SIZE=2>I apologize for the shameless plug. I work for Yahoo! ;)</FONT>
</P>
<P><FONT SIZE=2>> I plan to restructe the code a bit more. Id like to intodruce</FONT>
<BR><FONT SIZE=2>> files record.h/.c containing the code to acces the B-tree _Records_</FONT>
<BR><FONT SIZE=2>> and leave btree.h/.c for the structures _inside_ the Records. So</FONT>
<BR><FONT SIZE=2>> record would be something like:</FONT>
<BR><FONT SIZE=2>> typedef struct {</FONT>
<BR><FONT SIZE=2>> volume* vol;</FONT>
<BR><FONT SIZE=2>> hfsp_fork_raw* fork;</FONT>
<BR><FONT SIZE=2>> btree_nodes_desc node;</FONT>
<BR><FONT SIZE=2>> union {</FONT>
<BR><FONT SIZE=2>> btree_head head;</FONT>
<BR><FONT SIZE=2>> ... // not sure what Ill need here ...</FONT>
<BR><FONT SIZE=2>> } u;</FONT>
<BR><FONT SIZE=2>> } record; // or better node ?</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> Btree would become something like:</FONT>
<BR><FONT SIZE=2>> typedef struct {</FONT>
<BR><FONT SIZE=2>> volume* vol;</FONT>
<BR><FONT SIZE=2>> record rec;</FONT>
<BR><FONT SIZE=2>> UInt64 keyPos;// position of key in given record</FONT>
<BR><FONT SIZE=2>> hfsp_cat_key key;</FONT>
<BR><FONT SIZE=2>> hfsp_cat_extry entry; // file, folder or whatever</FONT>
<BR><FONT SIZE=2>> } btree;</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> Note that rec is not a pointer so that the btree can be used</FONT>
<BR><FONT SIZE=2>> to traverse different records (Like in your iterator code.)</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> I target a code Sequence as follows:</FONT>
<BR><FONT SIZE=2>> volume vol;</FONT>
<BR><FONT SIZE=2>> record catalog;</FONT>
<BR><FONT SIZE=2>> btreee tree;</FONT>
<BR><FONT SIZE=2>> UInt32 cnid;</FONT>
<BR><FONT SIZE=2>> volume_open(&vol, "/dev/hdx", HFSP_READONLY);</FONT>
<BR><FONT SIZE=2>> record_open_catalog(&catalog, &vol);</FONT>
<BR><FONT SIZE=2>> btree_open(&tree, &catalog);</FONT>
<BR><FONT SIZE=2>> icnid = btree_find_recursive(&tree, "MyFile.txt");</FONT>
<BR><FONT SIZE=2>> btree_copy_data(&tree, "/home/user/MyFile.txt");</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> This would be the core for a hfspcp tool ...</FONT>
</P>
<P><FONT SIZE=2>It seems more logical to me to have the following:</FONT>
</P>
<P><FONT SIZE=2>volume_open(&vol, device, flags);</FONT>
<BR><FONT SIZE=2>btree_open(&tree, &vol, CATALOG_CNID);</FONT>
<BR><FONT SIZE=2>record_init(&record, &tree);</FONT>
<BR><FONT SIZE=2>find_record(&record, searchkey);</FONT>
<BR><FONT SIZE=2>catrecord_copy_data(&record, outfile);</FONT>
</P>
<P><FONT SIZE=2>This seems more in line with the actual relationships of the</FONT>
<BR><FONT SIZE=2>on disk structures, whereas I would find the other way to</FONT>
<BR><FONT SIZE=2>be somewhat misleading. In this case, the record would need</FONT>
<BR><FONT SIZE=2>the tree rather than the fork, and the tree needs the fork.</FONT>
<BR><FONT SIZE=2>However, I could just be biased by the way I wrote it. :)</FONT>
</P>
<P><FONT SIZE=2>> Im writing this while going by train to my father 60s Birthday.</FONT>
<BR><FONT SIZE=2>> His brain is so full of cancer that he will not know this fact</FONT>
<BR><FONT SIZE=2>> anymore. Ill need some undisturbed time and X-Windows to make </FONT>
<BR><FONT SIZE=2>> these changes so I doubt I will do it this weekend, perhaps</FONT>
<BR><FONT SIZE=2>> I can spare some time next week.</FONT>
</P>
<P><FONT SIZE=2>I certainly understand not having the time to work on this stuff.</FONT>
<BR><FONT SIZE=2>Hopefully I'll get a chance to analyze a bit more and figure out</FONT>
<BR><FONT SIZE=2>where I'm wrong about how the filesystem works, since my code</FONT>
<BR><FONT SIZE=2>is still pretty broken.</FONT>
</P>
<P> <FONT SIZE=2>Brad Boyer</FONT>
<BR> <FONT SIZE=2>flar@pants.nu</FONT>
</P>
</TD>
</TR>
</TABLE>
</BODY>
</HTML>
